Featured Post

Welcome to the Forensic Multimedia Analysis blog (formerly the Forensic Photoshop blog). With the latest developments in the analysis of m...

Friday, October 28, 2011

Leahy Bill? Don't hold your breath

I normally don't wade into the turbulent world of politics, but I am going to today ... sort of. I attended a presentation recently, and (reading between the lines) I think that those who believe that the Leahy Bill (Criminal Justice and Forensic Science Reform Act) is dead in the water and nothing is going to happen are going to be in for a big surprise.

Meet The Executive Office of the President of the United States' Subcommittee on Forensic Science (SoFS).

According to their web site, "The purpose of the Subcommittee on Forensic Science (SoFS) is to advise and assist the Committee on Science (COS), the National Science and Technology Council (NSTC), and other coordination bodies of the Executive Office of the President on policies, procedures and plans related to forensic science in the national security, criminal justice, and medical examiner/coroner systems at the local, state and federal levels."

Ok. So what do they want to do with their office space? Good question. Here's what they say are their goals and objectives, "This Subcommittee was created to assess the practical challenges of implementing recommendations in the 2009 National Research Council (NRC) report entitled Strengthening Forensic Science in the United States: A Path Forward, and to advise the White House on how best to achieve the goals outlined in that report. The SoFS is charged with developing practical and timely approaches to enhancing the validity and reliability of the federal government’s forensic science activities. This includes ensuring that regional, state and local entities adopt best practices in forensic sciences, and facilitating a strong coordinated effort across federal agencies to identify and address important federal policy, program and budget matters."

With me so far. No opinion yet ... just straight reporting from their web site.

Begin opinion: Look at the purpose statement. Their purpose is to advise on forensic science as it relates to criminal justice (among other things) at the federal, state, and local levels. Their goals include advising the White House on how best to implement the recommendations contained in the NAS report. Now read this statement again, "This includes ensuring that regional, state and local entities adopt best practices in forensic sciences, and facilitating a strong coordinated effort across federal agencies to identify and address important federal policy, program and budget matters." How does the federal government assure compliance in other areas of policy? With money. Again, my opinion and my observations are thus - soon enough you'll be faced with a choice; implement the recommendations or face the loss of federal funds. No need for Congress to act. No need for the Leahy Bill. This will come by way of regulation, enforced with a simple tool (money). Don't believe me? How was "education reform" accomplished? By attaching funding strings to education bills like NCLB. Remember, in a general sense, what the federal government funds, it controls.

Please don't get me wrong. Best Practices are necessary. Consensus standards are a good thing. Everyone in our business should be working from an SOP. Things like quality assurance/control, peer-review, training, and so on are great. How many police labs accredit their DME forensic functions? Not many. How many would be ready to accredit them soon? Not many. How many agencies could withstand the loss of some/all federal funds if their DME forensic functions were not under their main accreditation umbrella? Not many.

In doing my part to help things go right, I'm working with LEVA to create an accreditation program for DME labs. It will serve as an alternative for those agencies who are not ASCLD/LAB accredited and are performing DME forensics. More on this later.

Again, while my commentary may be a bit political in tone, it's more of a "here's the way I see it based on my recent experiences." I'd be interested to hear how you see this working, or how you see this affecting your agency.

Enjoy.

Thursday, October 27, 2011

Copy Move Forgery Detection

Here's the abstract from an interesting paper that a reader alerted me to: "Abstract—One of the principal problems in image forensics is determining if a particular image is authentic or not. This can be a crucial task when images are used as basic evidence to influence judgment like, for example, in a court of law. To carry out such forensic analysis, various technological instruments have been developed in the literature. In this paper the problem of detecting if an image has been forged is investigated; in particular, attention has been paid to the case in which an area of an image is copied and then pasted onto another zone to create a duplication or to cancel something that was awkward. Generally, to adapt the image patch to the new context a geometric transformation is needed. To detect such modifications, a novel methodology based on Scale Invariant Features Transform (SIFT) is proposed. Such a method allows both to understand if a copy-move attack has occurred and, furthermore, to recover the geometric transformation used to perform cloning. Extensive experimental results are presented to confirm that the technique is able to precisely individuate the altered area and, in addition, to estimate the geometric transformation parameters with high reliability. The method also deals with multiple cloning."

Before you run out and look for the program to buy to perform these functions, relax. There isn't one that lets you load a single CCTV still image and press the magic CSI button that runs the SIFT algorithms described in the paper. There may be one in the future, but there isn't one now. Many image scientists used programs like MATLAB to perform these types of experiments.

Enjoy.

Wednesday, October 26, 2011

Can Your Digital Images Withstand A Court Challenge?

This just in from the Digital Forensic Investigator news service: "... What most agencies fail to realize is that the lack of SOPs and a sound workflow means images submitted for court purposes may not survive if challenged by a knowledgeable attorney. These digital complexities have not yet been realized, so images taken by photographers will likely fail one of three very basic criteria:

  • The date and time setting in the camera is incorrect. 
  • The wrong lens focal length was used, resulting in an inaccurate depiction of the scene. 
  • Embedded information may indicate the image has been modified.
“These points were not an issue with film, but we do NOT live in a film world anymore. We now live in a digital age, and it changes everything we once believed about how photographs are treated in law enforcement,” says D. Eric Johnson, CEP (Certified Evidence Photographer instructor and retired First Lieutenant, Michigan State Police). A strong advocate of digital image integrity, education, and certification, he feels it is crucial for all who work in law enforcement to understand the following: “A ‘true and accurate representation’ is no longer the only qualifier—and is a court challenge waiting to happen. The only question is when and where ...”

... he goes on ...

So Ask Yourself: Can Your Images Withstand a Court Challenge?
"It is time for those taking crime and accident scene photographs to take a hard look at their procedures and consider the following facts:

  • The majority of cameras in the field have an incorrect date and time—check yours! 
  • Compact point-and-shoot cameras default to a wide angle setting, which will misrepresent the accuracy of the image—and most officers don’t even realize it. 
  • Digital single lens reflex cameras (DSLR) come with zoom lenses. There is only one position on that lens that will accurately represent the image—and most officers don’t know it. 
  • When images are downloaded to a computer, the simple act of rotating an image in order to view it will indicate it was “modified.” Will your court accept a modified image? 
Any one of these issues will likely result in the images being disallowed if challenged—yet guaranteeing digital image integrity is not that difficult. Providing proper training, policies, and having at least one CEP in your agency is a small investment compared to losing just one important case in court."

Click here to read the whole article.

Enjoy.

Source Range

Hey! Media Encoder is ignoring my ins/outs. What gives?

The little pull down menu called Source Range is defining what gets exported. Make sure that the setting matches what you want to do.

Enjoy.

Tuesday, October 25, 2011

Gamut mapping

The question came up during the conference as to differences in the size of the source and destination spaces. Here's an answer from Adobe's Jim King: "When the gamut of the source and destination colorspaces differ a "gamut mapping" needs to be performed. Those colors that cannot be represented in the destination need to be altered to colors that can be represented.

If one is working with a company logo or any other object where it is important to preserve the colors colorimetrically, then one will want to colorimetrically match source colors to matching colors in the destination. For those colors that have no match some "closest possible" match should be found. This type of gamut mapping is called "colorimetric".

Experience has shown that colorimetric gamut mapping of pictures can be improved upon and so a "perceptual" method of gamut mapping is preferred for pictures. Usually these mappings make changes even to the colors that can be matched. The human visual system is more sensitive to relative color differences than it is to absolute differences so changing the mapping to preserve those differences yields better looking results.

When producing a business graphic or other schematic material, it may be more important to have the best solid saturated colors that a device can produce than to have an accurate color that the device produces in a poorly rendered way. So for these kinds of objects one uses a "saturated" gamut mapping.

There is a correspondence between gamut mapping methods and what has come to be called "rendering intents". More about rendering intents a little later.

The ICC has agreed to a basic set of four of rendering intents shown here. For the perceptual and saturated intents considerable latitude is allowed in the interpretation and implementation since no one has demonstrated a single best method to do the gamut mapping."

Enjoy.

Monday, October 24, 2011

AES 46th Conference on Audio Forensics

Dates: June 14-16, 2012 Location: Denver, CO, USA Chair: Jeff M. Smith If you wish to be contacted with conference updates please click here. Authors, the Call for Papers is now available.

Friday, October 21, 2011

Stanley Cup riot video analysis a tiresome task

This just in from the Vancouver Sun: "The young woman with light brown hair sports a delirious grin as she uses a hockey stick wrapped in a Canucks flag to smash the rest of a 15-foot store window, after game 7 of the Stanley Cup Final in June.

More than four months later, her image now sits on the desk of Special Const. Jevon Vaessen as he begins identifying her amid riot video footage which, if screened continuously, would last 208 days.

“It’s lot like searching in, say, a Microsoft document or something,” said Vaessen, one of the Integrated Riot Investigation Team’s four full-time forensic video analysts. “You’re looking for the key words.”

In the search fields of his computer program, he types in his “target’s” upper clothing colour and length (blue and short), lower clothing colour and length (dark and long).

About 20 clips show up with people in frames matching those descriptions.

Vaessen then inputs another search field with the woman’s hair colour (light and long). Ten clips show up.

“It’s like a Google search, you start searching and say, ‘Is this what I need?’” said Vaessen, who is on loan to the VPD from Port Moody police department.

Vaessen will screen each clip until he can obtain an image of the woman from which an investigator can get an identification. Within one frame, analysts could target as many as 10 rioters, identified by red dots.

“There are some clips with so many targets,” Vaessen said. “You’ll see red dots side-by-side-by-side.”

To date, 91 suspects have turned themselves in or been positively identified.

The video searching is made possible by the meta-tags that a team of 60 international forensic video experts at the University of Indianapolis applied to the riot clips over a two-week period in September.

The university’s lab was established in 2007 and houses FBI training courses, but this was the first time an active investigation used the lab. The analysts working on the riot case were paid by their individual law enforcement agencies, but IRIT footed the bill for their room and board.

The analysts amalgamate footage shot from video cameras, smartphones and digital cameras that was in dozens of different formats into a single file type.

Then Vaessen, his fellow IRIT analysts and their American and British counterparts went through each frame and flagged them with meta-tags denoting location and describing the people and events.

These 15,000 different tags include multiple camera angles of the same suspects committing the same crimes.

Some critics say digging for this amount of evidence is overkill and bogging down the riot probe, but lead investigator Insp. Les Yeo said IRIT is legally required to collect all possible evidence before charges are laid.

“We have to follow that procedure,” Yeo said. “If we don’t follow that procedure we’re going to get bad case law, acquittals, not guilty or no charges.

“It’s actually something that is forensic evidence, just like a fingerprint,” Yeo said of the evidence produced from the video analysis. “And, just like DNA, we have to forensically get it in as evidence to the court.”

Yeo said he understands the public’s frustration with the pace of the investigation.

“I totally get it. I, too, am not happy with the fact that it takes this long to get the charges forwarded [to Crown prosecutors],” Yeo said. “However, working within the system and with having worked within the system for 30 years I do know what is required ...”

Continue reading the article by clicking here.

Enjoy.

Thursday, October 20, 2011

Cloud computing and criminal justice not compatible?

This just in from Govtech.com: "... Levin said there is one issue outstanding that must be resolved for Google Apps to be fully compliant with the FBI’s Criminal Justice Information Services (CJIS) requirements for data storage and security. She didn’t divulge publicly what the unresolved problem is.

“The real issue here is the fact that the policies related to a lot of different areas in the government are not matching the technologies that are coming out,” Levin said. “That is the core issue: The criminal justice requirements were never written with cloud computing in mind.”

It’s taken a long time and some interpretation, she said, for all those involved — the city, Google, the FBI and federal government — to figure out how to make Google Apps compliant with the federal government’s security requirements.

In a press statement, CSC said it’s continuing to work with Los Angeles to meet these e-mail requirements. A statement from Mayor Antonio Villaraigosa’s press secretary said the city and Google continue to work together to resolve issues related to new Department of Justice e-mail requirements. “Discussions between the city and Google over specific requirements are ongoing and it would be inappropriate to issue further comment at this time," Villaraigosa’s office said.

Levin added that Google Apps isn’t the only product of its kind that hasn’t yet achieved full CJIS compliance. For example, Microsoft’s Office 365 hosted e-mail solution also fails to meet CJIS requirements, she claimed.

This week’s developments represent the latest curveball in the two-year-old story that has made Levin and the Los Angeles Information Technology Agency a battleground for cloud computing providers that are trying to stake claim to the public-sector market.

In 2009, Google hailed Los Angeles as a trailblazer when officials agreed to move the e-mail and productivity applications of all city employees — including the LAPD — onto Google Apps. Some observers thought the move was bold because it meant that public data would be stored off-site in Google’s data centers. Levin admitted in a later interview with Government Technology that the city didn’t intend to become the first large U.S. city to attempt an enterprise rollout of Google Apps. The city, she said, had misinterpreted the extent of Washington, D.C. government’s earlier implementation of the solution.

Migration of the LAPD’s e-mail system proved problematic from the beginning, as officials from the police department told Levin and the L.A. City Council two years ago that there was concern about storing and securing data with a private company. Levin thought she had assuaged the outcry after receiving assurances that Google Apps would be fully compliant with the FBI and Department of Justice regulations. But progress has been slower than anticipated.

In a press statement Tuesday, Consumer Watchdog said the outcome of L.A.’s contract with Google has “serious implications” outside of Los Angeles for other government agencies that have adopted Google Apps. In a letter of its own addressed to Villaraigosa, the group asked the city to “disclose immediately the extent to which Google has failed to comply” with its contractual obligations. The deal has amounted to “broken promises and missed deadlines,” the group claimed ..."

Read the whole article by clicking here.

Enjoy.

Wednesday, October 19, 2011

Links of note

For the students in the Intermediate Forensic Photoshop class, here's a few helpful links.

  • Stairstep Interpolation - Fred Miranda's SI Pro plug-in can be found here.
  • PercepTool 2 can be found here. You can also find George's book there too.
  • My book, Forensic Photoshop, can be found here.
Remember, your workflow is your reliable and repeatable pattern of activity. Make sure to document it and stick with it. If you need to change it, make a note of why it changed and when.

Of course, if you need some help, feel free to contact me.

Tuesday, October 18, 2011

The future of deblur

From the Photoshop blog: "We added Smart Sharpen in CS2, but deblur technology wasn’t mature enough yet for Photoshop and it’s been nagging me ever since. Given the nature of the heavy computation needed, the technology is really dependent on the evolution of the hardware, which provides a more powerful CPU and GPU for us to leverage.”

Challenges with Deblur

However, there is still quite a bit of development left to do before this feature is ready for prime time. Although some of these early demos will wow audiences, there is a lot more to blur than meets the eye. For instance, there are algorithms that estimate where blurs occur in an image and determine what type they are, and others that then reconstruct the image.

The before image [above] has a blur caused by camera shake. The after image shows the type of magic that can occur when the right algorithm is applied using Jue’s new prototype.

The tricky part is when an image has more than one kind of blur, which occurs in most images. Current deblur technology can’t solve for different blur types occurring in different parts of a single image, or on top of one another. For example, if you photograph a person running and also shake the camera when you press the shutter, the runner will be blurry because he is moving and the whole image might have some blur due to the camera shake. If an image has other issues like the noise you often get from camera phones, or if it was taken in low light, the algorithms might identify the wrong parts of an image as blurry, and thus add artifacts in the deblur process that actually make it look worse.

Strong edges in an image help the technology estimate the type of blur. The image below shows the same algorithm run on the image above without the benefit of strong edges. You see that it fails in this case ..."

Click here to read the whole story.

Enjoy.

Monday, October 17, 2011

LEVA 2011 Conference in Coeur d'Alene

I will be teaching the Intermediate Forensic Photoshop classes this week in Coeur d'Alene, Id at the 2011 LEVA conference. The dates are the 18th and 20th. I look forward to seeing y'all there.

Thursday, October 13, 2011

Gamut compression

Here's a nice explanation of what happens when you use an old printer in a modern workflow, from Adobe's Jim King: "Since the most notorious color gamut compressions involve reducing the total number of colors that can be represented going back to the original values is usually impossible. What has been lost is lost.

So it is best to delay any steps where gamut compression can occur to avoid reducing the gamut of the data we have. It must be done for output to gamut reduced devices but it is best to reduce your data in this way only in the last step."

In other words, work your workflow. Save your file for discovery. Then target for your old printer and print your exhibits.

Enjoy.

Tuesday, October 11, 2011

New Tool for Police, the Video Camera, and New Legal Issues to Go With It

From the NY Times: "When a man was fatally shot by a police officer on a street in Oakland, Calif., late last month, the shooting was captured by a video camera.

But the video was not taken by an alert pedestrian with an iPhone. It was recorded by a device clipped onto the police officer’s chest.

The Oakland Police Department is one of hundreds of law enforcement agencies that are trying out the body-mounted video cameras, using them to document arrests, traffic stops and even more significant encounters, like officer-involved shootings.

The cameras, legal experts say, are the latest addition in a world where everyone is increasingly watching everyone else.

The police already record illegal left turns and ignored stop signs using cameras mounted on the dashboards of cruisers — evidence displayed vividly on video screens in courtrooms, sometimes to the chagrin of drivers who have just insisted they did no such thing.

Surveillance cameras watch for shoplifters and potential terrorists. And ever since a bystander recorded Los Angeles police officers beating Rodney King in 1991, video has been used by witnesses or suspects to record what they believe to be misconduct or inappropriate behavior by the police — a practice that has proliferated with the advent of smart phones.

The ubiquity of video in police encounters — some of it promptly uploaded onto YouTube — is creating new frontiers for judges and lawmakers, who must sort out the issues raised by the new technologies.

Courts in several states are considering cases where citizens who videotaped the police have been charged with violating wiretapping or eavesdropping statutes, prosecution that civil rights lawyers say violates First Amendment rights ..."

Continue reading this article by clicking here.

Enjoy.

Monday, October 10, 2011

Diving into NVIDIA GPU’s and what they mean for Premiere Pro

From Adobe's Dennis Radeke: "Trying to quantify the GPU in a Premiere Pro for the editing and pro community, it is a big deal. It is something that I’ve wanted to do for some time, but I had no idea how complicated and time consuming this would be!

I went into this endeavor thinking that I would clearly delineate between different Quadro level cards and along the way understand what each one offered in the way of performance. I was methodical in setting up my system and in trying to create real-world tests that would emphasize what the GPU brought to the table.

While I did learn a lot (which I will pass on to you in due course), the experience and tests in some cases were not always as clear-cut as I had expected ..."

Click here to keep reading this article.

Enjoy.

Thursday, October 6, 2011

Remembering Steve Jobs

Via John Nack, the Photoshop Team remembers Steve Jobs.

For me, I remember the gift that Woz and Steve gave me and my college crew. The move from the IBM XT to the Mac Classic brought new opportunities and allowed us to graphically realize the products of our imagination ... not to mention earning extra bucks by making party flyers ... Aside from the brief detour that was the G3 (yes, I cursed them for making me learn NT), I've been a Mac guy ever since.

Wednesday, October 5, 2011

Another battle of the experts leads to a new trial for Wisconsin man

From the State Bar of Wisconsin's Joe Forward: "Brian Avery, convicted on armed robbery charges in the 1990s and sentenced to 30 years in prison, will get a new trial thanks to digital imaging technology not available at the time of his trial, a Wisconsin appeals court has ruled.

In 2007, Avery filed a motion for post-conviction relief based on newly discovered evidence. The new evidence was derived from a new method of digitally enhancing video surveillance, known as Video Image Stabilization and Registration, developed by NASA. It allows examiners to ascertain the height of individuals in surveillance videos, among other things.

A jury saw “grainy” video surveillance tapes of the robberies and pinned Avery on two separate armed robberies, weighing other evidence at trial.

But a digital enhancement, known as a photogrammetric analysis, using the new technology showed the robbery suspect was actually several inches shorter than Avery.

Two experts testified, one for the state and one for the defense, about the precise height of the individual (thought to be Avery) captured on video during one of the robberies. The defense expert testified the suspect could not be six feet, three inches (6’3”) tall, Avery’s height.

The state’s expert testified the suspect appeared shorter, between 5’11” and 6’1 ½", but noted that other variables made it possible the suspect was 6’3”. The circuit court denied Avery’s motion for a new trial after weighing the expert testimony.

A new trial is warranted, Avery argued, because this new evidence, along with other evidence, could place a reasonable doubt in the jury’s mind about his involvement in the robberies. In State v. Avery, 2010AP1952 (Oct. 4, 2011), the District I Court of Appeals agreed.

“By concluding that [the defense expert’s] opinions were ‘not reliable enough’ to entitle Avery to a new trial, the trial court gave one opinion from a credible witness greater weight than a competing opinion from a different credible witness” wrote Judge Joan Kessler. “A trial court applies an incorrect test when it weighs competing credible evidence.”

The trial court should have determined whether there was a reasonable probability that new evidence would create doubt about the defendant’s guilt, the appeals court explained.

“Simply put, if the jury believes the new evidence from Avery’s expert, then it would conclude that Avery could not be the man in the video,” Judge Kessler wrote.

The appeals court also ruled that Avery was entitled to a new trial under Wis. Stat. section 752.35, which gives a court of appeals discretionary reversal authority if the real controversy has not been fully tried. The real controversy, Avery’s involvement, was not fully tried, the court explained, because the jury did not hear the photogrammetric evidence.

The circuit court had ruled that new evidence would not destroy the state’s case against Avery. But the appeals court noted that “[n]o Wisconsin case interpreting [section 752.35] requires that the defendant’s new evidence” totally destroy the prosecutions theory ..."

Click here to read the story on the Wisconsin State Bar's web site. Enjoy.

Tuesday, October 4, 2011

Hurray for the Chief

You've got to give it up for the Chief of Vancouver's PD. He gets it ...

"The urge to rush to justice is basic and visceral and it is easy to see why many people want speedy arrests for the Vancouver rioters. Despite these external pressures, the Vancouver Police Department-led Integrated Riot Investigation Team will not cut corners and rush cases to court in order to appease those who want fast “justice.”

Many have expressed frustration with the Canadian justice system especially in light of the swift prosecution of the British rioters. But Canadian police and Crown attorneys operate under our laws, not foreign laws. The Integrated Riot Investigation Team (IRIT) and our prosecutors have set out to do the best possible under the current Canadian laws and court precedents. We work within the system as it stands.

The main hurdle we have been dealing with during our investigation is the unprecedented quantity of multi-format video evidence. We received more than 1,600 hours of such video. It would take years to process using conventional methods. We have taken this huge volume of digital evidence to a special lab at the University of Indianapolis, the world’s leader in processing digital evidence. Video experts from around the world will create a single-format video stream. We will be able to catalogue each riot participant and locate each case of vandalism and looting that the person committed. And then we’ll present that evidence to the court ..."

Continue reading this excellent piece by clicking here.

As it turns out, there were more than 5000 hours of video in this complex case. Reflecting back on my role in the IRIT, I was yet again the "Adobe guy." It seems that no one who was present in the opening days of work was proficient in performing batch rename functions on large volumes of data. Being the "Adobe guy," I knew that this task was easy using Adobe Bridge.

There are certainly other tools that could perform the function, but Adobe Bridge makes the job fast and easy. Using Adobe Bridge, I was able to append case and evidence ID information to the file names of the volumes of evidence we received for processing in an afternoon. In this way, those tasked with further processing did not have to perform this task manually, which often leads to mistakes.

I am very grateful to the VPD Chief for his sending a formal commendation and certificate to my chain of command for my work on the IRIT. 

Enjoy.

Monday, October 3, 2011

TWAIN is dead ... long live WIA Support

TWAIN's been dead for a while now. Some folks, just upgrading machines or Photoshop, are finding that the venerable TWAIN interface is gone. What to do?

File>Import>WIA Support is the new TWAIN (well, it's old ... but let's not get that technical).

Here's the scoop from Microsoft:
"Windows Image Acquisition (WIA) enables imaging programs, such as Microsoft Picture It! 2000, Kodak Imaging, or Adobe Photoshop, to communicate with imaging devices such as digital cameras and scanners. WIA supports digital still cameras and both low-end and high-end scanners; it also enables you to retrieve still images from IEEE 1394-based digital video (DV) camcorders and Universal Serial Bus (USB) Web cameras."

Enjoy.

Saturday, October 1, 2011

Can You Identify Any of These Suspects?

Wow ... a week in Indianapolis and I need a vacation. What a week it's been. I'm amazed ut and humbled at the talent that LEVA assembled to assist the IRIT and Vancouver's PD. I'm tired ... but I'm so glad that I was a part of it. Many thanks to VPD for their hospitality and to the IRIT for their generosity. Eddie, my row-mate, did some amazing things capturing the un-capturable video. Steve proved that even Newf's can learn to use the text/tagging feature in Avid (though you don't have to like it). U Indy's cafeteria proved that you can have chicken and chips every day for lunch (the gluten free menu).

From the front of the room, to the back, and on both shifts, there was just an absolute display of talent. At the front of the room was the inimitable Grant, conducting the symphony. I don't think that we'd have done it any other way. I'm tired ... but I loved every minute of it. BTW Scott, I got a guy too ...

Enjoy.