Very enlightening, wilhelm, and put in layman's terms I understand. I like the thought of scans being "hearsay". Also Ann-Willow's "road & cars" analogy.

CD pirate--

You are correct, but that is why I said "can" instead of "do" when describing how scan speeds themselves can distort the facts. That "5%" and the "hardly ever" must be included in a statement because it does sometimes happen. People tend to have complete faith in a scan as if their drives were making no contribution to the results.

High-speed drives introduce their own mechanical flaws into the reading (and writing) of discs at high speeds. Errors that appear as the result of radial noise may actually be due to drive flaws rather than a poor moulding process, but that can only be verified by test equipment that independently checks the disc's groove geometry. In those cases, a poor drive can ruin the reputation of a good disc at high speeds.

But of course . Keep in mind that temperature can also influence your results. A scan made shortly after switching on the PC in a cool room will likely be different from one made after the PC has been running for hours in a warm room. Give the drive a couple minutes to cool down between scans (especially if hi-speed) as well.

I'm not sure if it's better to start a new thread with this question. Maybe Arachne/DrageMester can advise .

Yes, please. A new thread in the Blank Media forum is preferred for such a question.

I would like to add that, using a CAV reading method for scanning can introduce its own problems, because of signal processing that takes place at the front end of the drive.

Very interesting post. I actually enjoyed reading that

Dee's post before mine is very interesting also, and hopefully will prompt further discussion

I hope I didn't either confuse people or oversimplify things in describing the parity error matrix. That was a secondary point trying to explain "parity inner" from "parity outer" designations. The parity numbers are control information added to the data in order to isolate and correct a bit that appears to be wrong or unreadable. This seemed to be a fundamental question Ann-Willow had in interpreting scan results and profiles.

wilhelm, enlightened members won't be confused by your post/-s.

But to make it easy and simple, sometime we have to take shortcuts and scan at higher speeds on behalf of accuracy.

Keep on posting, I enjoy it.

@wilhelm - no confusion here either. I felt things were very well explained.

I do agree somewhat with your post, and btw, I do think your posts are nice and clear as other people have stated

Now back to the high speed scanning hehe. I think a high speed scan can ruin a good discs reputation @ high speeds but that is only if you use high speeds to scan And you must agree that simply scanning @ 4x can make a bad disc look good. I've seen it many times before. And truly bad discs can be made to look even better once again if you do not include jitter scanning.

Almost every single "great" burn I have will scan well @ slow speeds and still fairly well @ 16x. Even T02, which scans with higher errors, still scan extremely acceptably @ high speed in both my liteons.

I scan @ both high and low speeds because if you don't do both, I believe you are leaving yourself open to a higher chance of letting a bad disc through. Take this example - I burn a MCC004 on my LG4167B @ 12x. Just using a PIE/PIF scan @ 4x, no jitter, all my drives show this disc is a very nice burn. Introduce jitter, it shows a large amount of fluctuation and goes extremely high in some spots. This disc exhibits playback problem @ certain points in my stand alone player.

Just using 4x scanning or even 8x scanning without jitter and without high speed scanning, I will have a disc that appears perfectly burnt, well within all error specs, but won't work flawlessly in my player. I then introduce 16x scanning - the disc shows extremely elevated errors in the spots where jitter spikes up. The 16x scan successfully finds the problem due to difficulty in reading high jitter @ high speeds. 4x is simply far too slow to show problems when jitter is high.

I think of jitter like this - you have two people hand writing on a piece of paper. One writes very messy. One writes very clearly - they both make about the same amount of errors in spelling (PIE/PIF). The messy writer and clear writer are read without problems by the person reading, he is reading slowly though so the spelling mistakes are easily seen and correct by him. The boss comes in and tells the reader he is going to slow and wants him to proof read the stuff twice as fast! The clearly written paper is read without problems again, he can still read through it fast despite some spelling errors because it's written so clearly, the messy written paper is difficult to read and the proof reader encounters more problems reading it faster and can't make out words as well. This is how I like to think high speed scanning of discs with high jitter works.

You also say that radial noise can affect a disc when reading @ high speed and make the drive show more errors. I disagree somewhat. Why do really high quality discs like A grade CMC MAG E01 or MCC02RG20 show zero problems when read @ high speed? They show zero problems on all my drives. Including my most fussy scanners. I tend to think these discs are somehow superior in some way since they scan so perfect in every single drive, every single speed, and never display playback problems in a wide range of stand alone DVD players. If your drive was somehow notorious for creating errors @ high speed due to radial noise it caused, surely it would do this for all discs? Why would some discs scan so beautifully and TRT with absolutely zero problems if radial noise caused by the drive was occurring? I often to see a large correlation between discs with crappy jitter levels, and rises in errors with 16x scanning. This could also be due to other out of spec errors on the disc that are unmeasurable by home based drives. But once again, that is a problem with the disc, not the drive.

My conclusion is extremely good burns will scan with low errors @ all speeds in different drives. The drive cannot be blamed for creating errors due to radial noise when multiple different drives are used and all show problems @ high speeds. 4x scanning without jitter just by itself is leaving out too much of puzzle and is far too basic to be considered a valid measurement of quality. It may be reliable and extremely consistent but in a powerful new drive, it's just too slow and forgiving to expose enough problems.

The problems of high speed scanning cannot be ignored. IMO, problems shown on a disc scanned at 16x that didn't show on a disc scanned at 4x accompanied by a perfect transfer rate test, in most cases has more to do with the inherent problems of scanning at 16x using a CAV reading method.

Has anyone taken note of the amount of air that is moved by a disc spinning at 16x? It stands to reason that the unclamped outer edge of the disc must be oscillating to a certain degree, as wilhelm has already explained very well.

There is also this trend now of trying, to explain things that cannot be explained, into neat little packages and labeling them as fixed or explained. For example. Saying the problem was shown on a 16x scan proves that a 16x scan is better for showing up a problem on a disc that passed at 4x scanning speed. And people haven't taken into consideration that 16x scanning creates its own problems and they may only be seeing a flaw in the reading drive rather than a problem with the burned disc.

IMO, to much blame is hurled in the direction of a burning drive here on CDF. It never seems to enter into anyones head that maybe the reading drive is flawed.

Well, reading drive being flawed is something people should consider @ any scan speed. There's loads of people being harsh critics of burnt discs after using NEC/Pioneer drives also.

I'll sum this up in points here for everyone to understand.

PIE/PIF scanning is most useful for comparing burn speeds of media when:

A: Same media is used from the same spindle
B: Same drive is used to burn and only speed is changed
C: Same firmware is used
D: Same drives/speeds are used to scan the burnt discs (you need consistency with tests)

There's probably more than that but you all probably get the gist of it.

Now a good use of high speed scanning -

A: Disc burnt @12x - good 4x scan - bad 16x scan - bad playback in standalone
B: Playback issues @ spot with bad jitter and high errors in 16x scan
B: Disc burnt @8x - good scans @ high & low speeds, lower jitter - no playback issues
C: All discs used from same spindle, same writer used same firmware etc
D: Results are repeatable when repeating the whole burn/scan process

You could say right there that there's something wrong with the disc burnt @ 12x. High speed scanning has shown something that correlates well with the problematic playback. 4x scanning only without jitter simply cannot detect these issues. TRT with a decent drive also can't detect this. Accompanying 4x/8x scans with a high speed scan is a good way to gain some extra info, possibly uncover potential flaws. And just so everyone knows, I've witnessed 16x scanning show many flawed discs that scanned well @ 4x. It's not just my 4167B @ 12x MCC004 - no one trick pony's here.

With all scanning, you have to look @ it not so harshly. Remembering that even some discs that seem out of spec or not so pretty may work perfectly fine in real world operation. They may also scan differently, better or worse in a different reader.

I don't think 16x scanning is superior at all, I also don't think 4x scanning is either. IMO using both and more than one drive, different makes/models also is the smartest and most thorough way to go about it.


Btw - it seems Dragemester is creating a thread about this topic. I'll continue posting there and if you want, Drage, you can move some of my posts to it if we're off topic here and or you post is/isn't open for discussion

Yep, I've been wanting to show my best examples for a while but never got around to it until now.

I decided to use my prerogative ( ) as a Blank Media moderator to create a new thread about it, even though there are at least 2-3 threads on this topic already.

My excuse is that due to the large number if scans in my examples, it's better to have the examples in the beginning of a new thread instead of at the end of an older thread.

For anyone interested, here's the new thread (now open for discussion):

Why high-speed scanning is good / Why high-speed scanning is bad

This seemed to be the most appropriate thread to reopen this topic: Why is there not more attention paid to high PIE levels/totals in a DQ scan?

Quite a few times I've seen others' scans showing high PIE but low PIF, and so because the disc plays and TRT's well it is accepted as being fine. But isn't the PIE scan actually a "raw" reading of the disc's written pits -- before any error correction measures are taken (and the info is converted to useful data) -- and therefore a better indicator of the true quality of the burn?

Perhaps Jitter readings are so important because they relate closer to the raw write than the corrected one, and high speed scanning can be beneficial because it shows more uncorrected errors?

Lower PIE levels/totals in the writing = less work for the player's error correction system during reading = better compatibility with players -- particularly those with less-capable error correction systems.

I understand that a DQ scan is not actually reading the data off the disc as a TRT does, but it is still reading the pits burned on the media. So I would then assume high PIE levels/totals are a sign of a bad write -- an early warning that if the player's error correction setup isn't up to the task there will be playback problems...



Which leads to the next point: Isn't there genuine importance in creating discs with low PIE DQ scan results?

I know some people get obsessed with obtaining a perfect scan while others pay very little attention to that parameter of disc write quality evaluation. I know that the scan numbers can vary by a considerable amount depending on the drive's manufacturer and chipset. But the scan can still give useful info regarding the raw readability of a disc. Regardless, isn't a high PIE result in terms of levels/peaks (say over 80) and totals (say over 100,000 on a >4G write) a sure sign that the burner/strat/MID pairing was not a good match?

Even though a modern player's error correction setup should be able to handle fairly high PIE (I have discs with sustained 400+ PIE levels and multi-million totals that play and TRT fine at the moment), how can one be sure with that with the passage of time the readability won't degrade faster than a disc with low PIE ?

I'm certainly not trying to attain a perfect scan; I consider any DVD write with PIE peaks <80, PIE totals <100,000, PIF quality >90 and PIF total <800 to be a good enough burn to trust with important data for up to five years. But I do worry when I see higher PIE scans...

Look at how much these results vary. Remember this is the exact same disc. One scan in a BenQ 1620 and one in a BenQDW1800, which is a liteon with different FW.

Massive difference in almost all aspects. These MCC03RG20 are not A grade IMO. They all seem to have some inbuilt dust at the edges of the discs and this shows in every single scan as the PIF gets more concentrated at the end.

I believe this is a good working disc though. I trust it. The whole idea of degradation - having lower errors at the start makes it degrade slower than a poorer burn is ridiculous and dumb if you think about it IMO.

If a disc degrades, and degrades plenty in a short amount of time, what difference will the initial errors make?

Think of it like this: Pits and lands start to go funny, reflectivity dives in random spots, regardless of whether the disc is written well, it will quickly become an equally bad burn as a poorer burn because of the degradation.

If you have one disc with 10,000PIE and 200PIF totals and another with 600,000PIE and 2000PIF, what difference will it make if both discs then get degraded and they then both have 8,010,000PIE and 25,200 PIF compared to 8,600,000 PIE and 27,000PIF? Know what I'm saying?

I don't know if slow degradation is even possible or if anyone has even documented it. IMO, once a disc starts it won't stop and things can only get worse with a possible snowball effect.

To answer your question - PIE is far easier to correct than PIF is. Blocks of PIF can make the disc unreadable for some devices. PIE is also more variable between devices whereas PIF is always more similar between reliable scanners. You can have one 16x model of Liteon, and a 20x model, if you scan @ the same speed (probably 4x), there's far more chance that the PIF will be similar than the PIE. Just look at my results above also, similar PIF but PIE is ridiculously different.

Thanks for responding. I'm really trying to get this whole business of home scanning sorted out in a logical way, to determine if it has suffient merit to base media buying decisions -- or not.

Here's where I'm coming from: all a "write" on an optical disc is, is an incredibly large number of pits "burned" (or any name you care to use) in the dye, in a pattern determined by the logic of the data stream and the write drive to manipulate the write laser. The accuracy of the write is determined by how compatible the MID, write strat, and physical ability of the write laser are to follow the strat in terms of laser power, focus and positional accuracy.

My belief is that the readability of a write is directly related to the conditions in the previous paragraph, and high PIE levels/totals are the first sign of "bad" write. Why wouldn't the normal chronic deterioration of a disc with abnormally high PIE numbers be worse than one with lower PIE levels? At some point the PIE amounts will be too high for error correction to work well.

But my core questions remain:

1) Isn't a PIE scan basically a raw error measurement before error correction is applied, and therefore the more accurate measurement of the quality of a write?

2) Regardless of which scanning drive is used (except whether it is a 1EEC or 8ECC type, which could give different but still usable results) high PIE amounts indicate the scanner is having difficulty reading a disc -- a sign of a bad write or failing scanner. Correct?

I have two 20x Liteys and an older 12x NEC and they all agree to within about 5-10% of each other on all aspects of a DQ scan regarding PIE levels, patterns/slopes, and totals. If I did the reg hack to permit scanning with my 203B's they would probably also be in that range. So it sure seems that a number of my drives share a fairly similar accounting regarding the PIE levels when scanning the same disc.

To me it seems important to match MID/burner/strat in order to generate a burn that the majority of scanners agree on that have reasonably low PIE amounts.