I get asked regularly about how I go about testing my DVD media so here is an insight on how I do it with standard PC equipment.

I usually accept samples to test in full retail tubs or spindles - this way I can see if there could have been any transit damage due to poor packaging.

I examine the suface of the discs for imperfections, paying particluar attention to their outer rim. I have never found one completely perfect one so don’t be alarmed when you don't either. Examine the underside of the disc for irregularities in the dye and scratches on the disc. The scratches do not reflect the quality of the disc but have a bearing on the results you may get. Packaging of the discs has improved and recently there have been little or no problems in this area.

Next we need to test the whole of the disc so an ideal movie should be one that has a total file size of just under 4.38gigs which will burn the disc to its limits at the outer edge. For test purposes I always burn at the maximum speed rating for that particular disc (i.e. I don't initially try to burn at 8x if it is a 4x rated disc).

After the burn I examine the burn surface for irregularities and prior to checking the usability do a file transfer scan with DVDTestPro using a Toshiba SD-M1712 DVD-ROM drive that has had the read speed limit restrictions lifted in the firmware, Because of the high speed reading now available with this drive I have to limit the read speed to 12x which I find gives me a good indication as to the readability of the media. 16x read gives a very erratic graph with most media. 8x read gives a smooth graph with most good media. The graph for 12x can be smooth or rough depending on the media. I always use the same DVD-ROM drive at 12x read limit to give an accurate comparison between different discs.


The following read speed graphs show the same disc in the same drive read at 16x, 12x, and 8x to indicate my point.



16x speed


12x speed


8x speed

After a read speed I run the KProbe test with my Liteon DVD-ROM drive. I am looking here for the lowest possible reading. Ideally the PI (Parity Inner) errors should not rise above 50 and the PO (Parity Outer) errors above 5 (with an average of less than 10 for the PI and less than 1 for the PO). These, of course, are the ideal figures I am looking for and the acceptable specifications are a lot higher.

An acceptable PI would have no large areas on the disc exceeding 280. Do not worry too much about high single spikes that exceed 280. An acceptable PO would have no large areas on the disc exceeding 32 PO (actually PI uncorrectable) errors, do not worry too much about high single spikes that exceed 32.

I do not profess to know a lot about the actual PI/PO errors other than that they are errors on the disc that are corrected by information on the disc. Any one interested in knowing all about the DVD specs can download THIS DOCUMENT and look forward to heavy reading.

Now that I have burnt the disc and checked the readability it is necessary to see if it works in as many DVD players as is possible. My most fussy DVD player is an old Mustek and I use this player to first check the disc as it is prone to freezing in playback with the lower grade of media.

Firstly I check how long it takes to load the intro or menu and listen for any unusual noises from the player drive as it starts to read the disc.

Playback the main film and watch the first few minutes for any signs of pixilation or the film freezing.

Fast forward and reverse at all speeds to check for continuity and smoothness.

Chapter skip forward and reverse to check if it finds the chapter points quickly and easily.

Play back the last 10 minutes of the disc and scrutinise the output, the reason for this is that the outer edge of the disc is the most likely area to have any defects.

I then try the disc in as many players as is possible (I currently have about 10) and the disc should play in all of them.

I have a few more players than most and the discs I test have to play in my worst players before I will give it the A OK. I have some of the newer players and they will play disks I have thrown out because of playback errors, so a good player is not ideal for testing media quality.

I personally only burn Movies which are for backup purposes so my tests are biased this way but it is possible burn data or PS2 games for the test, the most important thing is to try and make the content on your test disc as near to a full disc as is possible and test it to its limits on as many devices as you can.

The type movies to burn for test purposes need specific layout of the VOB structure which can affect the playback of a poor quality disc and for my tests I use three different types

1: A complete movie only with just the intro & title followed by the film which covers the full size of the disc this will give an easy test of chapter skips fast forward and reverse and playback consistency towards the end of the disc

2: A multi movie enabling to select different movies on specific parts of the disc and the ability to select the last movie to scrutinize the outer area of the disc

3: (This is one that will not play if the outer area is difficult to read or has errors) Some movies especially Disney ones have the habit of having the intro or warning on the last VOB on the disc, followed by various screens that appear to be random VOBs on the disc, this means that when the movie starts the laser has to jump to the rim of the disc to play it before going back to the title and then play the movie. Any problems with the disc will show up by the way it starts up it can start with a title intro or still warning and then go blank whilst it searches for the next part of the intro, sometimes it will find the intro and play it with errors and others it will remain blank for ages and then go to the start of the film skipping the intro entirely, the worst case it will just remain blank and lock up with disc error. If the intro starts normal and the movie starts to play you have the makings of a good burn.

I think the DVD producers do this deliberately to give us difficulty in making backups of their movies and in difficult cases remastering makes the disc more playable (but that’s another subject)

Since I started writing this article Nic Wilson of DVDInfoPro has commenced development on his next release v3 and one of the main features he has added is PI/PO & Jitter testing in conjunction with the NuTech DVD burners, and the provisional appearances impressed me so much I went out and purchased a NU DVDRW DDW-082 so that I could try this out. I am amazed at the detail produced by the graphs and although there is a noticeable similarity in the graphs. The detail & accuracy are by far superior to the KProbe tests and the actual scaling is much better.

I have now built up a dedicated DVD media Test Rig so that I can produce better comparisons.



This Test Rig has all my test drives mounted externally to the case so that if I decide to change any of them it will be easy.

It consists of the NU burner, LITEON LTG166 reader

TOSHIBA SD-M1612 & TOSHIBA SD-M1712 readers in addition to 2 removable HD bays

My PI/PO/Jitter test are done on the NU drive and the speed tests are done on the TOSHIBA 1712 drive which has had the read speed limitations removed and then externally restricted to 12x speed which I find is the best speed for me to compare discs.

This is an example of the type of graph to expect from the DVDInfoPro3



To make sure my tests produce a fair comparison between medias I never burn a copy on the fly I always burn either a disc image or a title set and for the scans I make sure the burn is as near to a full disc as is possible.

My scans are all done with the same drives and any unusual results I rescan to make sure it is correct, I like to burn at least 10 discs for each particular media and test them for consistency before I can make a final conclusion and any single discs that come out exceptionally good or bad are noted but not counted in my final evaluation.

Well that’s how I test my DVD's with standard equipment, If I was to go in for the more professional approach and had the money to spare.

I would go for the AudioDev CATS system.



For Further Info & Resources Visit:
Flash's website: http://flashman.rpc1.org

Added:  Tuesday, May 11, 2004
Reviewer:  Flash

hits: 13167