5 ms·
Complete speculation here and I have no idea if it's actually possible, but couldn't you take a high resolution photograph of the DVD and then use software to t
by onion2k 8y ago
Complete speculation here and I have no idea if it's actually possible, but couldn't you take a high resolution photograph of the DVD and then use software to turn the image data back in to DVD data? Maybe with a couple of different colored light sources at obtuse angles to the disc? It's obviously not a practical solution but it'd be really fast if you could make it work.
It's definitely possible with phonograph records.
- Dylan16807 8y agoConsider that you'd be trying to photograph 100 billion bits of pit and land. To capture it properly a terapixel image might work. Nothing can do that in one photo, or short series of photos.
- beojan 8y agoNo. DVDs are much higher density than phonograph records.
- U2EF1 8y ago[citation needed]
- dikei 8y agoJust a rough calculation: Each pit on the DVD represents 1 bit, but need at least 1 pixel to stored. You need 8 bit to store a pixel. So you have write amplification of at least 8 when taking picture of DVD.
- onion2k 8y agoThat's why I suggested using a couple of light sources - you could take an image that's lower resolution than 1 pixel per bit and still resolve the image data to the DVD data based on the color. Eg 10 would be a different color to 01. It's not even remotely practical but I thought it was an interesting idea. Judging by the way the votes are yoyo'ing up and down, about half of HN readers agree. :)
- newnewpdro 8y agoI'd assumed some optical magnification would be required in combination with multiple photos per disc. Then software would have to sort out the mess.
- Rotten194 8y agoThink about it -- a vinyl can hold ~22 minutes of music per side, and a DVD-R can hold ~120 minutes of video per side, in a much smaller area.
- TomMarius 8y agoCome on, you can calculate that without a calculator.
- laumars 8y agoYou don't even need to do the maths. Just looking at a vinyl would give the game away since you can see grooves with your naked eye (as a DJ in an earlier life, we used to use that as a visual clue for when to queue up the next record, swap the basslines, etc). Not to mention how albums would often be spread over two records where as the same album could fit on a single audio CD (let alone DVD). I'm guessing the GP has never owned any vinyl. I could forgive someone questioning the density of records if he's not familiar with the tech.
- onion2k 8y agoI'm guessing the GP has never owned any vinyl. I could forgive someone questioning the density of records if he's not familiar with the tech. I'm old enough to know about vinyl. That's not really the point though - the principle is the same, just with much higher information density. With a suitably high resolution camera, or set of cameras, or video, and some serious magnification it should be possible to photograph a DVD and play back the data from the image. I find that quite interesting, or at least entertaining, to think about. I guess the HN readers who're downvoting the post don't. Something to consider - technically we can already do it - it's how DVD players work, albeit with a single coherent light source and only reading one bit at a time.
- laumars 8y ago> I'm old enough to know about vinyl. That's not really the point though I was talking about user U2EF1 (the "citation needed" comment). I didn't have any particular issue with your comment aside the impracticality of it with current consumer hardware (which I'll address below). But as a concept it's an interesting point. > the principle is the same, just with much higher information density. The problem is in the detail. Even with gramophone records, you'd get a very low success rate with a consumer camera. Plus you'd need multiple macro pictures and a precise way to stich them together. At which point it would be quicker to play the record while recording it in Audacity (or similar) in real time. So you're talking a less than 1x record speed and lower success rates to boot; and that's just low density vinyl. When talking about DVDs you'd need to improve the operation by several orders of magnitude in terms of accuracy and resolution. So I think your comment was interesting from a conceptual point but we're a long long long way off having that level of detail in consumer photographic devices. > Something to consider - technically we can already do it - it's how DVD players work, albeit with a single coherent light source and only reading one bit at a time. I think it's a little disingenuous having a laser reading binary reflections sequentially compared to a digital sensor detecting literally billions of analogue reflections (because a camera isn't just detecting the existence or absence of light) in parallel and then forming a precise sequence of digital bits from that. The technologies are completely different, the scales are completely different, the concept is completely different. They're just not equatable. > I find that quite interesting, or at least entertaining, to think about. I guess the HN readers who're downvoting the post don't. It wasn't me who downvoted you but if it's any consolation, I've gotten downvoted for factually accurate comments before (let alone impractical suggestions). You just have to remind yourself that it happens occasionally and usually the positive outweighs the negative. :)
- cge 8y agoIf I recall correctly, even for vinyl, "high resolution" in the context of demonstrations that had decently high quality audio reproduction meant scanning over the record with a microscope, with resolution significantly beyond what would be possible with a normal camera.
- newnewpdro 8y agoThis sounds like a fun DIY project to attempt.
- hamandcheese 8y agoMaybe if you shot a laser at the DVD and imaged 1 bit at a time. And then maybe you could rotate the disc while you do this, so you can get a lot of bits in rapid succession.
- seszett 8y agoJust think about the amount of data stored on a DVD. It's about 4 GB, which means that if you got a perfect image of just the data of the DVD (one pixel imaging each bit) the size of the picture would also be 4 GB. That means, as far as I understand, a 4 gigapixel image AT BEST, assuming you were able to get every single pixel to capture one bit of data on the DVD. In reality, since DVDs are circular and images are square you're looking at images that would be more like 6 gigapixel if my math is correct. Today, available cameras are at best 50-megapixels, which means you need at the very minimum about 100 photos of each DVD, but in reality, there is no way you can ever get one bit per pixel, because to begin with the bits are not aligned in rows unlike the camera's pixels, so you're most probably looking at ten times that. It's an interesting idea but I'm pretty sure it's completely unpractical. Edit and I realised I counted 4 GB, but I should actually have used 32 Gb as a starting number, as what matters here are the bits, not the bytes, so add another 8x to the amount of photos you need.
- social_quotient 8y agoUnhelpful to GP. Pied Piper could do it. Would use their “middle out” compression algorithm - anyone with a 5.2 Weissman score could do this pretty easily.
- slavik81 8y agoThis is actually an interesting idea in some ways, but there's such an enormous difference in scale between the problem and your solution that it's completely impossible. Like the bit with the guy trying to jump across the English Channel: https://youtu.be/Qvk2wNWmB20?t=47s https://youtu.be/Qvk2wNWmB20?t=47s