Jump to content

badreg

Member
  • Posts

    965
  • Joined

  • Last visited

Everything posted by badreg

  1. Just post them on this thread. Again, I don't know your market, so I have no idea if something is a good price or not. But there are plenty of Europeans on this forum who will be able to chime in. The build above is great on paper, for your budget, so you can use that as a guideline for what type of specs you should be looking for.
  2. 1.29v is not going to get it done. Probably less than 5% of 8700k's will be able to do 5.0Ghz on that voltage. Try 1.35v as a starting point and make sure that LLC is set to an appropriate level for your board.
  3. I can't. I'm not in Ireland so I have no idea what your local market is like. In general, prebuilts can be more expensive, but if they're on sale for Cyber Monday, that will often make up the price difference. The real downside to prebuilts is not price. It's the quality of the non-marquee components like PSU, motherboard and case. If you do get a prebuilt, make sure that the retailer lists the exact model numbers of these components so that you know what you are getting.
  4. Have you considered a prebuilt? If they are on sale, prebuilts will be close in price or cheaper than what you can assemble on your own, especially in the current market. The build listed above is literally impossible to purchase at MSRP. The street price of a 3070 is a lot more than its €500 MSRP. If you don't have a lot of experience, going to prebuilt route is probably the best way to get something quickly and save you from a lot of potential pitfalls.
  5. As pointed out above, the 1000-digit number needs to be declared as a string, and you need to iterate over the string. This is a theme for the more difficult problems later on, where you will need to perform operations involving multiple large numbers that cannot be represented as an int. A couple other problems with your code: There's no need to sort. You just keep track of the largest product and print that out at the end. You seem to be skipping the multiplication of first index of each substring by only multiplying when b > 0. a = "7316717653133062491922511967442657474235534919493496983520312774506326239578318016984801869478851843858615607891129494954595017379583319528532088055111254069874715852386305071569329096329522744304355766896648950445244523161731856403098711121722383113622298934233803081353362766142828064444866452387493035890729629049156044077239071381051585930796086670172427121883998797908792274921901699720888093776657273330010533678812202354218097512545405947522435258490771167055601360483958644670632441572215539753697817977846174064955149290862569321978468622482839722413756570560574902614079729686524145351004748216637048440319989000889524345065854122758866688116427171479924442928230863465674813919123162824586178664583591245665294765456828489128831426076900422421902267105562632111110937054421750694165896040807198403850962455444362981230987879927244284909188845801561660979191338754992005240636899125607176060588611646710940507754100225698315520005593572972571636269561882670428252483600823257530420752963450" x = 0 for i in range(len(a) - 13): temp = 1 for j in range(13): temp *= int(a[i + j]) if temp > x: x = temp print(x)
  6. Assuming that your 650W is a quality unit, you have plenty of headroom with your current setup. Adding a 1030 for 2d output is going to draw an additional ~10W, which is insignificant.
  7. You can't. The bill of materials is close to $500 for Sony. Consoles are sold at a loss for the first several years after launch.
  8. HDR is for HDR video content creation and consumption. Viewing HDR photos on a monitor isn't really a thing. In general, "HDR photography" refers to images that are tone mapped down to 8-bits. OLED displays are great for consumption, but they are poor choices if color accuracy is the priority. IPS panels have terrible black levels compared to OLED, but color accuracy is the reason why almost all high-end professional displays are IPS. Most people who view your images will not be viewing on OLED displays, and you have no idea how they would look on different displays unless you process and proof on a color accurate display. Finally, HDR does not improve black levels. It increases the peak brightness, which increases the contrast ratio. But for photo editing, you are generally going to calibrate to ~120cd/m2 and 750:1 contrast ratio is sufficient. The only factors at the $500 price point that you should consider (besides size and resolution) are gamut coverage and how well it will calibrate to sRGB and Adobe RGB (preferably 99%+ of both and a max Delta E below 3).
  9. Why is HDR a consideration if this is used for photo editing? A wide gamut (Adobe RGB) should be far ahead on the feature list than HDR.
  10. No, Adobe RGB is a different color space from sRGB. They are not related to each other. It just happens that Adobe RGB includes all of sRGB, but this is not always the case. For example, DCI-P3 is an emerging standard and is roughly the same size as Adobe RGB, but there are colors that are included in Adobe RGB that are not in DCI-P3, and there are colors that are in DCI-P3 but not in Adobe RGB. In essence, color spaces are just definitions of what the red, green, blue primaries are. Your display needs to be physically capable of outputting those primaries to display the entire color space. Calibration merely adjusts the brightness values of the subpixels so that the colors are displayed accurately. Calibration cannot make a display output a color that it was not capable of before. If you are doing this professionally, spending some time to learn a bit about color management will be a good investment.
  11. I'm not quite sure what you mean by a "plain" monitor and calibrating it "yourself". You may be misunderstanding how calibration works. First, in order to calibrate a display, you need a colorimeter in order to create a custom ICM profile specific to your display. The i1Display Pro that I recommended is one such device. Calibration is not something that you can do by eye. Second, if you are targeting the Adobe RGB color space, you need a display that is capable of covering the entire Adobe RGB gamut. Cheap displays typically cover only the sRGB gamut. Calibrating a display does not widen its gamut. That said, the majority of 10-bit displays are wide gamut and will cover 98%+ of Adobe RGB, so if 10-bit is one of your requirements, you should be good with regards to gamut. In terms of features that you may want to consider: True 10-bit vs 8-bit + FRC. Cheaper "10-bit" displays are almost always 8-bit + FRC. You need to decide if you require true 10-bit. Hardware calibration. This allows you to calibrate the internal LUT of the display. With software calibration, you are changing the signal that the GPU is outputting, and this results in a slight loss of both bit depth and contrast ratio. With hardware calibration, the processing takes place on the display, so it will be calibrated across different inputs and does not result in a loss of bit depth.
  12. The X-Rite i1Display Pro (and whatever the Datacolor equivalent is) should be what you are looking for. This is the cheapest device that can do hardware calibration on supported displays, and you shouldn't be using anything cheaper for professional work. However, it should go without saying that a colorimeter can only calibrate a display. It can't fix a faulty one or increase the gamut of a display that does not otherwise support Adobe RGB. If your display is dying, a colorimeter is not really the first thing that you should be thinking about.
  13. This is a side by side of 100% crops of RAW images from an iPhone 11 and a Sony A7R4. Is there a visual difference between the two? Sure, but the Sony is a 61MP sensor paired with a $1,500 lens. Yet this "tiny sensor" comes enough to a "real camera" that the difference is meaningless for most people. And this is compared to a current $4000 camera. Against a $600 camera from 2011 with a $200 lens, the iPhone would be ahead. If you are interested in reading more, this is the article that the image came from: https://blog.halide.cam/inside-the-iphone-11-camera-part-1-a-completely-new-camera-28ea5d091071 And just to give more context to what I am saying: I am a commercial photographer, and I have more real world experience with pushing cameras to their limits than most people here. Is it possible to squeeze more performance out of a dedicated camera than a phone? Of course, it is, but that requires a certain amount of skill, experience and post-production. But even then, no amount of skill will overcome the advancements in sensor technology in the last 8-9 years. You simply can't push a base ISO RAW image from a 600D by more than +2.00 EV before the noise becomes overwhelming, while modern iPhones will do that with ease.
  14. An image search on Google shows that ©2016 is what it should say. https://www.techspot.com/products/processors/amd-athlon-200ge.188774/
  15. The way that dynamic range is measured in a lab (at least the way that DXOMARK does it) is a valid indicator of real world performance. It is expressed as the ratio between the highest luminance gray and the lowest luminance gray that has a signal-to-noise ratio above 1 that a sensor can capture. Effectively, the lower the noise floor, the higher dynamic range a sensor has. In the real world, this has a direct effect on how far the exposure slider can be dragged to the right before the shadow areas fall apart. With a low dynamic range sensor like a 2011 Canon sensor, a ISO 100 image falls apart from massive amounts of noise and banding after +2.00 EV. Current high-end sensors will do +5.00 EV with minimal artifacting at base ISO. This is how dynamic range is determined in the real world and it closely matches what the lab measurements are. And yet again, you fall back to anecdotal, subjective claims that have no evidence. Just because you think that something looks way better doesn't make it so. When you make a blanket statement that you can always tell if an image was shot on a phone or a camera (especially when you have not done an ABX test), all you are doing is showing your hubris and an unwillingness to accept facts and data.
  16. You know that nothing stops you from capturing in RAW with a phone, right? The sensor in your T5 has a dynamic range of around 11 stops at base ISO. The iPhone 12 Pro has a dynamic range of 14 stops, which trades blows with modern full frame and medium format sensors that hit close to 15 stops. A modern flagship phone sensor is miles ahead of a mid-range APS-C sensor from 2011. It is not even close.
  17. 4.6Ghz is the maximum 1-core boost clock. That's why it says "up to". All core boost clock is 4.3Ghz. Since you have a Z board, you can enable "Sync All Cores", which will change the boost table to 4.6Ghz all cores. Or just dial in a manual overclock. Nearly every 9600k will do 5 0Ghz if you have adequate cooling.
  18. Resolution and dynamic range? Definitely a modern phone. Form factor and interchangeable lenses? Dedicated camera.
  19. There's probably a checksum somewhere in the file. Look for other differences and try to reverse engineer how the checksum works.
  20. I never said that the 30% figure has anything to do with warranty. It doesn't. Manufacturer warranties are based on both time and TBW. Your Evo is still at 100% because your host writes are probably less than 10x of your drive capacity. CrystalDiskInfo only looks at the average number of P/E cycles of each block to determine the estimated remaining life. It doesn't care how old the drive is.
  21. The 30% health is taken from the value of "Average Block-Erase Count", which shows that each block has been through 427 P/E cycles. (*Assuming that the value is in decimal. If it is hex, then it is 1063 cycles.) This is roughly in line with the 250TB of host writes that your 525GB drive has done. For context, the stated endurance for a 500GB MX500 is 180TBW for warranty purposes. You are way past that on your drive. On the other hand, the vast majority of TLC drives will continue to work well beyond 1000 P/E cycles. The takeaway here is this: don't expect your drive to die any time soon, but don't be surprised if it does. The advice above is correct. Backup anything critical data that you have, but you can continue to use the drive normally other than that.
  22. It's a buffer overflow. When your function tries to write to words[20], it is actually writing to the memory address that contains the value of alpha, so the value of alpha is overwritten. Change the declaration of words to words[MAX_WORDS + 1]. Edit: My original explanation is incorrect. words is not a std::string, but an array of std::string, which is zero index. It is still a buffer overflow, because of these two lines: numWords++; words[numWords] = filetxt; numWords needs to be incremented after each line is read, so these two lines need to be switched.
  23. Paste is not pre-applied on a Noctua cooler, but it will come with a tube of Noctua paste. OEM paste is not the absolute best that you can buy, but it will be within a couple of degrees C of the best paste, so for most people, the included paste is just fine.
  24. Check your CPU clocks. The benchmark results show 100% CPU bound in the DX12 run. The GPU score is higher in DX12 compared to DX11, which is expected, but your CPU scores drop off a cliff, so that's where the problem is.
  25. Literally anything released in the past five years if all you are doing is displaying a video.
×